Research On Fire Location In Image Detection System

Fire is a kind of natural disaster which has strong sudden and destructive.Especially, in large special building, the special building structure and purpose increasesthe risk of fire, the speed of spread and the degree of harm. This make prevention andsave more difficult for fire task. The fire detection and intelligent monitor point savesystem is a fire edge tool. It can real-time monitoring scene, alarm and position fire, andcommand fire monitor for fire extinction.This article is already in the early fire detection system and accurate detection ofthe fire source premise, focus on the image of the large space building fire positioningtechnology. Based on primarily computer vision, digital image processing, opticalmeasurement and automatic control technology principle, space fire source point earlypositioning and fine positioning, and ultimately draw more accurate fire source point inthe three-dimensional coordinates of the large space building to fire guns as theextinguishing basis. Process include: this article positioning system has detected a fireimage feature extraction, feature matching, the calculation of the spatial locationprocess. Existing image-type fire artillery fire localization algorithm are the basic theparallel binocular or monocular scanning positioning accuracy is difficult to meetfire-fighting requirements in a large space within. Absorb ranges photoelectricmeasuring precision advantages, and to be improved to make it suitable for theconstruction of large space remote measurement, and to increase the positioning methodof the optical axis coincides together form a remote measuring system, in the proximitymeasurement retained parallel binocular positioning technology. supplement. Obtainedbased on this initial positioning data, directing the artillery fire crude aiming beamdirections of the initial positioning result, combined with the known observation information secondary camera calibration, and extracting the image of the beam spotimage with the fire source point compared. In accordance with the position of the sourceof fire and the remote data, remote data than the data of multi-sensor data fusion. Theamendments to the fire location information given the results of the fine positioning.The experiments show that the method ensures the accuracy of the fire sourcelocalization, and improve the speed and the efficiency of the three-dimensionalpositioning of the fire source. To satisfy the needs of the large space of the fire sourcelocalization.